05_palero Palero and Garcia 66 Science Diliman (January-June 2001) 13:1, 66-72 ABSTRACT The third-harmonic 355 nm output of a pulsed Nd-YAG laser is converted into UV, VIS, and NIR laser light by stimulated Raman scattering in high pressure hydrogen gas. Laser lines in the 223 to 309 nm and 416 to 865 nm spectral regions are generated by anti-Stokes and Stokes Raman shifting, respectively. Experimental results on the energy output, conversion efficiencies, spectral profile, and temporal behavior of the various Stokes and anti-Stokes Raman laser lines are presented. The first and second Stokes shifted wavelength with wavelengths of 416 nm and 503 nm yielded a maximum energy conversion efficiency of 60.4% and 58.8% , respectively. Key words: stimulated Raman scattering, Raman shifting, frequency conversion. Frequency Conversion of the 355 nm Nd:YAG Laser via Stimulated Raman Scattering in Hydrogen J.A. Palero* and W.O. Garcia Photonics Research Laboratory, National Institute of Physics College of Science, University of the Philippines Diliman, Quezon City 1101, Philippines E-mail: jonathan@nip.upd.edu.ph INTRODUCTION Stimulated Raman scattering (SRS) is a powerful technique for shifting the wavelength of laser radiation into another spectral region. It is capable of simultaneously generating multiple laser lines ranging from vacuum ultraviolet (VUV) to far infrared (FIR). Compared to other frequency conversion methods, it is simple, economical, robust, and capable of high conversion efficiency. SRS has been used in a wide range of applications, such as generation of short laser pulses, time-gated low light level imaging, and lidar measurements. In this work, the third harmonic 355 nm output of a pulsed Nd:YAG laser is shifted into the UV, VIS, and NIR region of 223 to 865 nm through SRS in high pressure hydrogen gas. The energy, spectral profile, and temporal behavior of the Rayleigh and SRS lines are measured as a function of the 355 nm laser excitation energy and hydrogen gas pressure. SRS conversion efficiencies are calculated and presented. Spontaneous and Stimulated Raman Scattering A simplified diagram of the spontaneous Raman effect is illustrated in Fig. 1. An incident photon (ωp) excites a molecule from the ground state to a virtual state. As the molecule relaxes to the first excited state, it emits a photon (ωS) with a frequency lower than the incident photon. This frequency down-conversion process (ωp > ωS) is called Stokes shifting. If the incident photon (ωp) is absorbed by a molecule initially in the first excited state, the emitted photon (ωAS) will have a higher frequency as the molecule relaxes to the ground state. The up- conversion of the photon frequency (ωp < ωAS) is *Corresponding author Frequency Conversion of the 355 nm Nd:YAG Laser 67 referred to as anti-Stokes shifting. One of the characteristic parameters in this phenomenon is the energy difference between the first excited state and the ground state called the Raman frequency of the medium. For hydrogen gas, this quantity is equal to 4155.25 cm-1 corresponding to the energy between the ground state and the first vibrational state. At a sufficiently high incident photon flux, a process known as stimulated Raman scattering (SRS) can occur. It is a nonlinear optical process arising from the interaction of an intense laser beam with a Raman active medium such as CaCO3, SiCl4, CH4, and H2. The stimulated Raman scattering process is shown in Fig. 2. In Fig. 2a-c, a pump photon ωp excites a molecule to the virtual state and interacts with a Stokes photon ωS to stimulate the emission of another Stokes photon ωS. Fig. 2d-f show the stimulated emission of an anti-Stokes photon by the interaction of an anti-Stokes photon and a molecule in a virtual state. In contrast to the incoherent radiation produced through the spontaneous Raman effect, SRS leads to the generation of coherent light that has the same wavelength and phase as the incident Fig. 2. The stimulated Raman scattering process. Fig. 1. A simplified diagram of the spontaneous Raman effect VIRTUAL STATE ( a ) ( b ) ( a ) ( b ) ( c ) ( d ) ( e ) ( f ) Palero and Garcia 68 photon. Furthermore, the SRS output is much more intense than the weakly scattered radiation from the spontaneous Raman effect. If the intensity of the first Stokes-shifted frequency (S1) is high enough, another SRS interaction can occur between S1 and the medium. This process lead to the generation of a second Stokes-shifted frequency (S2). Higher-order Stokes wavelengths are produced by this process known as SRS cascade. On the other hand, generation of higher order anti-Stokes wavelengths is also possible by a nonlinear process known as four- wave mixing (FWM). It involves the interaction of three waves to form a fourth wave. Fig. 3 shows the SRS cascade and FWM processes involved in the generation of Stokes and anti-Stokes wavelengths. Unlike the SRS process, the generation of radiation by FWM is sensitive to the phase mismatch between the wave vectors of the interacting waves. These leads to the low efficiency of the FWM process at high pressures where the phase mismatch is greatest. FWM also contribute to the generation of higher-order Stokes often to a greater extent than the SRS cascade. Table 1 shows the anti- Stokes and Stokes wavelengths generated by stimulated Raman scattering in hydrogen using a 355 nm laser excitation. Fig. 3. The SRS cascade and FWM processes involved in the generation of Stokes and anti-Stokes wavelengths In the case of SRS in gases, the plane-wave steady- state Raman gain coefficient gR is given by: gR = 2 λ2S ∆N dδ ____ ______ ___ , hcνS πc∆νR dΩ where λs is the Stokes wavelength (cm), νs is the Stokes or Raman frequency (cm-1), c is the speed of light (cm/ s), h is Planck’s constant (Js), ∆N is the difference in population between the initial and final states (cm-3), ∆vR is the Raman linewidth (cm -1), and dσ/dΩ is the differential cross section for Raman scattering (cm2/ sr). In the case of hydrogen, ∆vR = 11.2/p + 1.58p, where p is the gas pressure given in atm. An analysis ω P ωS1 ωS2 ω S1 ω S2 ωS3 ωP ωS1 ωAS1 ωP ωP ωS1 ωAS2 ωAS1 (a) (b) (c) (d) (e) Raman Order Wavelength AS4 222.9 AS3 245.8 AS2 273.8 AS1 309.0 S1 415.9 S2 502.9 S3 635.9 S4 864.5 Table 1. Anti-Stokes and Stokes wavelengths generated by stimulated Raman scattering in hydrogen using a 355 nm laser excitation. ( a ) ( b ) ( c ) ( d ) ( e ) Frequency Conversion of the 355 nm Nd:YAG Laser 69 of the gain coefficient as a function of pressure shows that it saturates for pressures greater than 20 atm. The Stokes intensity varies exponentially with the interaction length (z) as IS(z) = IS(O) exp (gRILz) where IL and IS are the incident laser intensity and the Stokes intensity. According to these equations, the Stokes intensity and the SRS conversion efficiency depend on laser beam intensity, laser beam geometry, and Raman active gas and its pressure. EXPERIMENTAL SETUP & METHODOLOGY Fig. 4 shows the schematic diagram of the Raman cell and the gas handling system. The Raman cell is constructed from stainless steel hollow cylinders that are welded together. Both ends of the cell are sealed with UV grade fused silica windows and o-rings. The cell is provided with inlet and outlet gas ports. A mechanical vacuum pump is used to evacuate the cell. Upon evacuation, ultra high purity hydrogen (99.9999% purity) is introduced into the cell from a gas cylinder through a gas regulator and shut-off valve. A mechanical pressure gauge monitors the hydrogen pressure. A schematic diagram of the experimental setup is shown in Fig. 5. The laser is a Q-switched linearly polarized Nd:YAG laser (Spectra Physics GCR-230-10) which operates at 10 Hz pulse repetition rate, a wavelength of 1064 nm, and 5–8 ns (FWHM) pulse width. 355 nm output is obtained by the use of potassium dideuterium phospate (KD*P) crystals. The laser beam is passed through a diaphragm (D) set at an aperture of 1.5-mm diameter. A Glan laser polarizer (GP) is used to vary the laser pulse energy. The laser beam then is focused into a Raman cell using a 50 cm lens (L1). A 30 cm lens (L2) placed at the exit port of the cell is used to collimate the Raman output into a Pellin-broca prism (PB) for separation into the Rayleigh, SRS Stokes, and anti-Stokes components. A pyroelectric detector (Molectron J25HR) is used to measure the pulse energies of the 355 nm excitation laser, Rayleigh, SRS Stokes, and anti-Stokes lines. The spectral profile of the excitation, Rayleigh, SRS Stokes, and anti-Stokes lines are measured with a computer- controlled monochromator (SPEX 1000M). The temporal behavior is monitored with an ultra-fast biplanar phototube (Hamamatsu R1328U-5) and a 500 MHz digitizing oscilloscope. RESULTS AND DISCUSSION The spectral profile of the Raman output at a maximum input pump energy of 6.5 mJ and hydrogen pressure of 80 psi is shown in Fig. 6. There are eight Raman-shifted laser lines: four Stokes and four anti-Stokes, covering the ultraviolet to the near-infrared wavelength region. Spectral line-widths (FWHM) ranging from 0.02 nm to 0.08 nm are observed. Stokes lines are stronger than Nd:YAG Laser (355 nm) Hydrogen Raman CellD GP L1 L2 PB Rayleigh, Stokes and anti-Stokes Output Fig. 5. Schematic diagram of the experimental setup Raman Cell Vacuum Pump H2 Pressure gauge Pressure gauge End caps vent Fig. 4. The schematic diagram of the Raman cell and the gas handling s y s t e m . Palero and Garcia 70 the anti-Stokes lines, and the intensity tends to weaken with increasing Raman order. Fig. 7 shows the temporal behavior of the Raman output at different 355 nm laser excitation energy. The input laser pulse, which has a Gaussian profile, is truncated at half its maximum to show the temporal behavior of the scattered pulses more clearly. At an excitation energy of 2.27 mJ, generation of the first (S1) and second Stokes (S2) are observed. The 355 nm Rayleigh scattering has a lower energy than the input laser energy and a distortion in the pulse shape is observed due to the energy conversion into S1 and S2. A delay time between the Rayleigh and S1 is evident in the figure due to the required pump threshold intensity to generate a Stokes pulse. S2 also propagates with a time delay in respect to S1 by the same reason, which also caused a distortion on the pulse shape of S1. The observed time delay in the generation of higher-order Stokes is a clear evidence of the SRS cascade process. At an input laser energy of 4.45 mJ, the increase in intensity of S1 and S2 is accompanied by the generation of S3 and AS1. Although S2 increases in intensity, distortion in its pulse shape becomes apparent because of the onset of S3. The first anti-Stokes AS1 is observed to propagate earlier than S2 or S3 which can be explained by the fact that anti-Stokes wavelengths are generated not by the SRS cascade but FWM. For instance, AS1 is produced by the interaction of S1 and two photons of the pump, and although it cannot propagate earlier than either S1 or the Rayleigh, it may be generated earlier than the higher order Stokes. An increase in intensity for all Raman-shifted pulses and the generation of a second anti-Stokes shifted pulse AS2 are observed at 6.45 mJ of input laser energy. At this point, distortion of the Stokes pulses which may be due to the increased efficiency of FWM at high laser intensity becomes very apparent. The parasitic behavior of the anti-Stokes pulses with the Stokes is due to the 0 5 10 15 R e lat iv e Time (ns) 0 0. 1 0. 2 0. 3 0. 4 0 5 10 15 N or m al iz e d I n te n si ty 0 5 1 0 15 L a s e r Ray le ig h S 1 S 2 S 3 A S1 A S2 E = 2. 2 7 mJ E = 4. 4 5 mJ E = 6. 4 5 mJ Relative Time (ns) N o rm al iz ed I n te n si ty Fig. 7. The temporal behavior of the Raman output at different 355 nm laser excitation energy R E S U LT S 0 0.04 0.08 0.12 0.16 In te ns ity ( a. u. ) Center Wavelength (nm) 223.0 245.8 273.8 309.0 415.9354.6 635.9502.9 864.5 AS4 AS3 AS2 AS1 Rayleigh S1 S2 S3 S4 Fig. 6. The spectral profile of the Raman output at a maximum input pump energy of 6.5 mJ and hydrogen pressure of 80 psi Center Wavelength (nm) In te n si ty ( au ) Frequency Conversion of the 355 nm Nd:YAG Laser 71 fact that generation of higher-order anti-Stokes through FWM is not restricted to the interaction of one set of waves. In fact, when n Stokes and anti-Stokes beams are present (total number, including the pump), there is a total of (n–1) (n–2)/2 distinct FWM processes among them. Fig. 8 shows the energy conversion efficiencies of the Stokes and anti-Stokes on the incident 355 nm laser energy at a hydrogen pressure of 80 psi. The threshold for the generation of the first Stokes (S1), S2, S3, and anti Stokes (AS1) occur at 2.0 mJ, 2.7 mJ, 3.9 mJ, and 2.3 mJ, respectively. The dependence of the Stokes and anti-Stokes conversion efficiencies with hydrogen pressure at 6.5 mJ of input laser energy is shown in Fig. 9. The conversion efficiency is defined as the ratio between the output energy and the laser input energy. The peak conversion efficiency of S1 of 38% occurs at 300 psi. For S2, the peak conversion efficiency of 55% is attained at 500 psi. The conversion efficiency of the anti-Stokes reaches a maximum at around 80 psi. Anti-Stokes and higher-order Stokes are generated primarily by FWM processes at low pressures during which the phase mismatch of the interacting waves is minimum. At pressures greater than 150 psi, the phase mismatch is increased such that FWM is reduced. Fig. 10 shows the optimum energy conversion efficiencies of the Stokes and anti-Stokes. CONCLUSIONS Stimulated Raman scattering in hydrogen is investigated as a method for generating laser radiation in the region of 223 to 865 nm (UV to NIR). A cascade process of 0 2 4 6 8 10 12 14 16 18 20 0 1 2 3 4 5 6 Input Pulse Energy (mJ) C on ve rs io n E ffi ci en cy ( % ) S1 S2 S3 S4 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 Input Pulse Energy (mJ) C on ve rs io n E ffi ci en cy ( % ) AS1 AS2 AS3 AS4 Stokes anti-Stokes Fig. 8. Energy conversion efficiencies of the Stokes and anti-Stokes on the incident 355 nm laser energy at 80 psi hydrogen pressure 0 10 20 30 40 50 60 70 0 150 300 450 600 Hydrogen Pressure (psi) C on ve rs io n E ffi ci en cy ( % ) S1 S2 S3 S4 0 1 2 3 4 5 6 0 150 300 450 600 Hydrogen Pressure (psi) C on ve rs io n E ffi ci en cy ( % ) AS1 AS2 AS3 AS4 Stokes anti-Stokes Fig. 9. The dependence of the Stokes and anti-Stokes conversion efficiencies with hydrogen pressure 60.4 58.8 9.8 2.2 5.9 2.4 1.1 0.6 0 10 20 30 40 50 60 70 C on ve rs io n E ffi ci en cy ( % ) S1 (700psi) S2 (700psi) S3 (150psi) S4 (90psi) AS1 (80psi) AS2 (80psi) AS3 (80psi) AS4 (70psi) Fig. 10. Optimum energy conversion efficiencies of the Stokes and anti-Stokes Input Pulse Energy (mJ) Input Pulse Energy (mJ) Hydrogen pressure (psi) Hydrogen pressure (psi) C o n ve rs io n E ff ic ie n cy C o n ve rs io n E ff ic ie n cy C o n ve rs io n E ff ic ie n cy ( % ) C o n ve rs io n E ff ic ie n cy ( % ) Stokes anti-StokesStokes anti-Stokes C o n ve rs io n E ff ic ie n cy ( % ) Palero and Garcia 72 generation of the Raman components are observed from the temporal behavior of the output. The energy conversion efficiencies of the Stokes and anti-Stokes are observed to increase with increasing energy and to exhibit saturation. Anti-Stokes and higher-order Stokes are observed to have optimum conversion efficiencies at low pressures due to FWM generation while S1 and S2 are optimum at high pressures. Energy conversion efficiencies of 60%, 59%, 10%, and 2% are measured for S1, S2, S3, and S4 Stokes wavelengths, respectively. On the other hand, 6%, 2%, 1%, and 0.6% conversion efficiencies are achieved for AS1, AS2, AS3, and AS4 anti-Stokes wavelengths, respectively. These results are comparable to the results of Papayannis et al. (1998) who reported the generation of UV and VIS laser light in H2 using a third harmonic 355 nm Nd:YAG laser pump with conversion efficiencies as high as 50% at the first Stokes shifted frequency. ACKNOWLEDGMENTS The authors acknowledge the assistance provided by Mr. Jose Omar S. Amistoso and Ms. Marian F. Baclayon, the support of the Philippine Department of Science and Technology through the Engineering and Science Education Project and the advise given by Dr. Caesar A. Saloma and Dr. Marlon H. Daza. REFERENCES Carnuth, W. & T. Trickl, 1994. A powerful eyesafe aerosol lidar: Application of stimulated Raman backscattering of 1.06 mm radiation. Rev. Sci. Instrum. 65: 3324-3330. Chu, Z., U.N. Singh, & T.D. Wilkerson, 1991. Multiple Stokes wavelength generation in H2, D2, and CH4 for lidar aerosol measurements. Appl. Opt. 30: 4350-4357. Duncan, M.D., R. Mahon, L.L. Tankersley, & J. Reintjes, 1992. Low-light-level, quantum-noise-limited amplification in a stimulated Raman amplifier. J. Opt. Soc. Am. B. 9: 2107-2121. Milton, M.J.T., G. Amcellet, A. Apituley, J. Bosenberg, W. Carnuth, F. Castagnoli, T. Trickl, H. Edner, L. Stefanutti, T. Schaberl, A. Sunesson, & C. Weitkamp, 1998. Raman-shifted laser sources suitable for differential-absorption lidar measurements of ozone in the troposphere. Appl. Phys. B. 66: 105-113. Papayannis, A.D., G.N. Tsikrikas, & A.A. Serafetinides, 1998. Generation of UV and VIS laser light by stimulated Raman scattering in H2, D2, and H2/He using a pulsed Nd:YAG laser at 355 nm. Appl. Phys. B. 67: 563-568. Yamada, T., T. Nakata, & F. Kannari, 1995. Subnanosecond Pulse Generation in the VUV by Dual- wavelength Pumped Stimulated Rotational Raman Scattering. IEEE J. Quantum Electron. 31: 1274-1284.